Similar effect sizes have been reported for the effects of conspiracy, pseudoscientific, and paranormal beliefs on authoritarian attitudes, which points to a conceptual problem at the heart of the conspiracy literature, namely lack of clarity as to what uniquely defines conspiracy beliefs and whether those unique elements contribute distinctly to authoritarian ideologies. To our knowledge, this is the first study to test empirically the predictive power of variance unique to each construct against covariance shared among these constructs when predicting authoritarian and anti-democratic attitudes. Online survey was administered to 314 participants in 2021 that included a battery of demographic and psychological measures. Hierarchical factor models were used to isolate unique variance from shared covariance among responses to items representing conspiracy, paranormal and pseudoscientific beliefs. Structural equation models were used to test their unique and shared effects on authoritarian and anti-democratic attitudes. We found that our combined measurement model of paranormal thinking, conspiracism, and pseudoscience exhibited exceptional model fit, and that each construct was strongly predictive of both SDO and RWA (r = 0.73-0.86). Once the shared covariance was partitioned into a higher order factor, the residual uniqueness in each first order factors was either negatively related or unrelated to authoritarian and anti-democratic attitudes. Moreover, the higher order factor explained the gross majority of variance in conspiracy (R2 = 0.81) paranormal (R2 = 0.81) and pseudoscientific (R2 = 0.95) beliefs and was a far stronger predictor (β = 0.85, p < 0.01) of anti-democratic attitudes than political partisanship (β = 0.17, p < 0.01). Strong partisan identifiers of both parties showed much higher romanticism scores than party moderates. When predicting authoritarian and anti-democratic attitudes, we found no empirically unique contributions of conspiracy beliefs. Instead, we found that a shared factor, representing a 'romantic' mindset was the main predictor of authoritarian and anti-democratic attitudes. This finding potentially explains failures of interventions in stopping the spread of misinformation and conspiracy theories. Conspiracy theory researchers should refocus on the shared features that conspiracy thinking has with other unwarranted epistemic beliefs to better understand how to halt the spread of misinformation, conspiracy thinking, anti-science attitudes, and even global authoritarianism.